New Trends in Informatics Research - NTIR

April 17, 2009 The University at Albany | State University of New York

This Year's Research Presenters

Siwei Lyu, Computer Science

Digital Image Forensics with Natural Image Statistics

Abstract:

Development of digital technology allows one to manipulate images in ways that were unimaginable ten years ago. As digital images are becoming a main source of information, fraudulent or tampered images are having a huge impact in media, law, politics and many other aspects of our lives. Thus, there is an urgent need for forensic techniques to authenticate digital images and detect tampering operations. Even though a tampered image may fool the human eye, its statistical properties are disturbed and different from those of the original image. Therefore, it is possible to use image statistics as effective forensic tools to detect "unnatural" tampering operations. The goals of this work are to find representations and models that can shed light on the characterizing statistical properties of natural images, and apply them to problems in digital image forensics.

Biography:

Dr. Siwei Lyu received his B.S. degree in Information Science in 1997 and his MS degree in Computer Science in 2000, both from Peking University (Beijing, China), and his Ph.D. degree in Computer Science from Dartmouth College in 2005. From 1998 to 2000, he part-time worked at the Founder Research and Development Center (Beijing, China), where he developed a script language for an interactive multi-media authoring system. From 2000 to 2001, he worked at Microsoft Research Asia (then Microsoft Research China) as an assistant researcher, where he worked on techniques to render and recognize digital inks, for which he holds two patents. From 2005 to 2008, he was a post-doctoral research associate at Center for Neurosciences, New York University. He is currently an assistant professor at Computer Science Department at SUNY Albany. His research interests are natural image statistics, digital image forensics, machine learning and computer vision.

 

Alexis Wichowski, Informatics

The Myth of Fragmentation: Assessing Political Information Online

Abstract:

Internet technology has provided people with unprecedented abilities to filter the information they encounter, leading many scholars to fear that people will be exposed to less diversity of perspectives and fragment into homogeneous interest groups. However, these fears rest on three assumptions: 1. online, opportunities for unintended encounters with a diversity of information are limited, 2. people primarily pursue narrow interests when consuming online content, and 3. that narrow pursuit of interests leads to less informed citizens.

This presentation describes dissertation research that will examine political information online to provide empirical data that tests these assumptions. This will be achieved using mixed data collection methods. First, screen captures of a sample of both popular and specialized websites will be coded and subject to content analysis. This data will reveal what diversity of political content is available to be encountered in both popular and specialized websites visited. Second, a sample of regular Internet users will be recruited to participate in a computer-lab browsing task and complete questionnaires addressing customization and informing practices. This data will provide answers about how narrowly people do pursue interests online, and how these pursuits relate to the process of becoming an informed citizen.

Biography:

Alexis Wichowski is a doctoral candidate in the Department of Informatics at the University at Albany. Her research focuses on online content and communication.

 

Jeong-Hyon Hwang, Computer Science

iFlow: An Internet-Scale Software System for Data-Intensive Computing

Abstract:

Advances in hardware manufacturing have enabled proliferation of various computing machines ranging from commodity servers and supercomputers to mobile devices and sensors. Furthermore, the pace of data production has been accelerating with the increasing assimilation of computation into everyday life. The confluence of these phenomena necessitates software systems that can efficiently process massive amounts of data by orchestrating a large number of heterogenous computing machines.

In this talk, I will introduce iFlow, our software system that aims to help users with limited experience in parallel and distributed computing easily develop Internet-scale data processing applications. In iFlow, users express the desired computation as a network of operators. Our off-the-shelf operators provide processing capabilities essential to typical data management applications. In addition, operators customized for specific applications can easily be developed and plugged into the system. Given a large number of computers around the world, iFlow strives to deploy operators in a most beneficial way and avoids service disruption by replicating operators and rerouting around network outages.

This talk will include a brief demonstration of the current iFlow system. To extend iFlow for broader types of problems, we are seeking collaboration opportunities with researchers in various domains, including Social Networks, Bioinformatics, Economics, and Earth Science.

Biography:

Jeong-Hyon Hwang is an assistant professor in the Department of Computer Science. He completed his Ph.D. in Computer Science at Brown University in 2008 and received M.S. and B.S. degrees in Computer Science from Brown University and Korea University, respectively. His research interests include databases, distributed systems, and networking. He has co-authored more than 20 research articles and filed a US patent application. He is a recipient of the 2005 SIGMOD best demonstration award and has industrial experience as a software development team manager and a software engineer.

 

Donald Robadue , Informatics

How digital is what divides us? Global networks of practice for coastal management

Abstract:

Learning is a core idea for building capacity and achieving tangible results to advance marine and coastal management in developing countries. Networks of practice are emerging to take advantage of information technology, with the notion that virtual communities might be a low cost way to share information and overcome some of the barriers to good governance and sustainable development by enabling leaders to become more effective. Fostering networks of practice across regions and continents is motivated by the need to reduce the profound isolation which practitioners feel as well as to build their personal knowledge and social capital. However, active networks are based on engaging in joint activities and building personal commitment and trust, and technology by itself cannot supply these essential ingredients. There is no substitute for face to face contact in the generation and transmission of knowledge that is most relevant to leaders working in the unique circumstances of every coastal ecosystem. A review of recent literature and experience reported by large international organizations provides the backdrop for an examination of the success and challenges of a network of Latin American coastal managers and a network of mariculture professionals in East Africa.

Biography:

Don Robadue is a PhD candidate in information science at UAlbany and a coastal management specialist at the University of Rhode Island Coastal Resources Center, where he has been since 1977. He currently works in Tanzania as an coastal planner; will be presenting the Center's work on Coastal Climate Change Adaptation at the upcoming World Ocean Conference in Manado, Indonesia; and is serving as an information specialist for a global project funded by USAID advocating an integrated approach to population, health and environment. He also leads an ongoing study of the utilization of Rhode Island's economically active waterfront.

 

Taewoo Nam , Public Administration and Policy

Potentials of the Internet as a Public Sphere: An Empirical Study of Political Activities Online

Abstract:

This study examines potentials of the Internet as a singular macro-level space of the public sphere, statistically analyzing three national surveys. With the focus on who participate in the online public sphere, what they do there, and how and why they do such political activity, it investigates whether the virtual sphere on the Internet has core components of the public sphere. Results from regression analysis and ANOVA indicate that the Internet fails both in contributing to the substantial improvement in absolute equality and in enhancing the level of political participation greatly. A set of independents including political efficacy on the Internet, civicness and political interest are strong predictors for the level of activeness in online political activity. Significant predictors are different for online vs. offline activity and thick vs. thin democracy activity. Notably, the effect of age on online political activity significantly differentiates more deliberativeness from less deliberativeness.

Biography:

Taewoo Nam is a second-year doctoral student in the Department of Public Administration and Policy, the University at Albany - SUNY. He is from Korea, and graduated from Korea University (Bachelor in Public Administration) and Indiana University (Master of Public Affairs). His research interests are e-government, e-democracy, and information technology policy. For publication, he is currently working on a cross-national comparison study of e-democracy and an empirical analysis on potentials of the virtual public sphere.

 

Namjoo Choi , Informatics

Non-Technical Users' Migration to Free/Libre Open Source Software and their Contributions to the Community

Abstract:

In the past several decades, Free/Libre Open Source Software (FLOSS) emerged as one of the most promising phenomenon in the software industry, and now it has evolved into a movement not only in the software industry but also in our society as a whole. Unlike the early stages in the FLOSS evolution when FLOSS was made mostly by, of and for developers, recent statistics clearly indicate that it has begun to embrace a rapidly rising number of non-technical users. It is assumed that the trend of this explosive rise in the number of non-technical FLOSS users whose population outnumbers that of technical developers lies in the locus of a paradigm shift, software commoditization.

Biography:

Namjoo Choi is a doctoral student in Informatics at the State University of New York at Albany. He has an MA in Telecommunication, Information Studies and Media from Michigan State University and a BA in Mass Communications from Yonsei University, Seoul, Korea. His research interests include information security, open source software, and decision support systems, e-government. His research work has been published in some referred journals and conference proceedings including Information Management and Computer Security, Journal of Information Technology and Politics, AMCIS and HICSS. He received the Outstanding Paper Award Winner at the Emerald Literati Network Awards for Excellence 2009 for the article entitled, Knowing is doing: An empirical validation of the relationship between managerial information security awareness and action, Information Management & Computer Security, 16(5), 484-501 and the best-paper runner-up at the 42nd Annual Hawaii International Conference on Systems Sciences (HICSS 42).

 

Jonathan Muckell , Informatics

Intelligent Sampling and Transmission of Vehicle Data

Abstract:

While efforts to produce and share private sector freight flow data are underway, barriers to sharing remain, including concerns with privacy, private sector data production costs and the quality and quantity of the final data outputs. New strategies for intelligent sampling and transmission for telematics hold great promise for future sharing opportunities and enhanced productivity for the freight community. Recent advances in telecommunications, electronic sensors and GPS technology are driving down the costs of on-board telematics systems. These devices monitors the location and status of sensors equipped on vehicles which is periodically send satellite/cellular messages to a centralized location. This technology is particularly popular for commercial fleet management by allowing in-transit monitoring of the supply-chain network. In particular this technology has applications in food/drug safety and for general efficiency reasons. Although there are numerous practical reasons for equipping telematics systems, organizations are often hesitant due to the long-run incremental cost of sending frequent satellite and/or cellular messages from the vehicle to a centralized location. Methodologies for restricting the number of transmissions, and thereby cost, either lose essential information or contain redundant data (e.g., many telematics devices will send messages only when the on-board sensors detect loading cargo, opening doors, etc.). This strategy introduces data gaps - the period of time between events when no information is received. Optimization of fleet operations and transportation planning application typically requires information that occurs during these gaps, which can be large, on the order of hours or even days. These periods of nonexistent or insignificant information prevents accurate data mining of transportation critical information, such as route prediction, travel time reliability and drive behavior. An alternative model of telematics communication sends information at regular, pre-defined time intervals (i.e. every two minutes). This strategy often sends redundant information, resulting in higher transmission costs.

Our approach uses an intelligent communication strategy that eliminates data gaps, while reducing the size and frequency of telematics messages. Using the Douglas-Peucker line fitting algorithm to compress the sampled points, the compression scheme minimizes the maximum error in the location reconstruction. Furthermore, the reconstruction is free from data gaps since it preserves both the spatial and temporal components of the GPS trace (i.e. when the vehicle was at each location). By reducing the frequency and size of telematics messages while preserving the information, the economic value of the telematics systems can be better applied for freight managements and transportation planning.

Biography:

Jonathan Muckell received his M.S. in Computer Systems & Engineering from Rensselaer Polytechnic Institute in 2008, and his B.S. in both Computer Science and Mathematics from St. Lawrence University in 2006. He is currently a Ph.D. candidate at the University at Albany studying Informatics, with a concentration in the utilization of geographic information systems for freight logistics and optimization. He has also developed numerous spatial technologies and published several papers as a member of the Defense Advanced Research Project Agency (DARPA) and National Geospatial Intelligence Agency (NGA) funded GeoStar Project. Jonathan is currently the newest member of the Visualization in Transportation Committee of the Transportation Research Board of the National Academies. His current research is often conducted in conjunction with the Pervasive Decisioning Systems Lab at the General Electric Global Research in Niskayuna NY, and more recently on a joint UAlbany - CUNY - Hunter College collaborative GPS project.

 

Matthew Jager , Informatics

Organizational Knowledge Management Through the Lens of the Community of Practice

Abstract:

The discipline of knowledge management (KM) has emerged in response to the growing emphasis on knowledge as an organizational resource and knowledge-related goals in management. Throughout its brief existence, however, the discipline has been criticized for failing to contribute any thing truly new to the greater practice of management, and risks being written off as another management fad. Specific criticisms of the discipline include a failure to coalesce around a framework or frameworks that can explain how the central concepts of knowledge management are related to each other, or even a commonly-accepted definition of knowledge, itself arguably the most important concept in the discipline.

A Community of Practice (CoP) has been defined as a [group] of people who share a concern or a passion for something they do and learn how to do it better as they interact regularly. As a mechanism of learning, the concept of the CoP has an obvious appeal to KM researchers and practitioners, and techniques intended to stimulate the activity of CoPs have become an accepted component of the KM practitioners toolkit. Some scholars have suggested, however, that a deeper relationship between organizational knowledge and CoPs may exist- that the CoP might be considered a fundamental unit of analysis, with respect to organizational knowledge. As such, the CoP might provide a site for analysis and theory-building that could give rise to the kind of unified framework, and understanding of knowledge in an organizational context, that the discipline of knowledge management currently lacks.

This study applies Social Network Analysis (SNA) to the membership of a large, complex organization, in order to reveal Communities of Practice. This Social Network Analysis is followed by ethnographic analysis observations and interviews of the members of the identified Communities of Practice, in order to generate theory concerning the relationships between people, processes, and technology in the creation, identification, sharing, and utilization of organizational knowledge.

Biography:

Matthew Jager is a PhD Candidate in the Information Science PhD Program in the College of Computing and Information at the University at Albany. He received his Bachelor's Degree in Physics from the University at Albany, and his Master's degree in Business Administration from the School of Business at the University at Albany. His research focuses on understanding organizational knowledge and developing the theoretical foundations of the discipline of Knowledge Management.

 

Xiaoai Ren , Informatics

Multiple Case Studies of Public Library Systems in New York State: Assessing Services to Provide to Member Libraries

Abstract:

Public libraries consolidation and merger are a long and heated topic among researchers and practitioners in the U.S. The same applies to public library systems in the U.S. In the U.S., the idea of the public library system evolved out of the need to extend and enhance the library services to the under-served and un-served populations and to provide services to member libraries based on the spirit of resource sharing and economies of scale. Public library systems receive funding from Federal government, state and local government as well to operate and coordinate the resource sharing activities among member libraries. In New York State in fall 2005, the New York Library Association (NYLA) Public Library Section (PLS) conducted a survey of public libraries to explore their attitudes toward the potential public library systems merger. Based on the survey, this study will explore the public library systems in New York State further.

Biography:

Xiaoai Ren. Ph.D. student at Department of informatics. Research interests: public library, public library systems in the United States, information policy and digital preservation.

 

Alex Chaucer , Informatics

Adding the time dimension: considerations for using and comparing digital geographic layers over time.

Abstract:

Traditionally maps capture geographic phenomena as a reflection of a specific date. There are many different means of geographic representation for the landscape, including topographic maps, aerial photos, parcel information, satellite-based land use classifications, census data, economic data and election data. In addition, the frequency of digital product updates has increased. This exploratory research examines how the time dimension may be visualized by looking at static layers in succession. What are the strength and weakness of looking at these digital products in dynamic succession over time in animation? What changes become evident when viewed temporally in succession? What are the potential application areas for this type of analysis and visualization? Our analysis will consider these questions, as well as some of the platform considerations for creation and distribution. These considerations may be of interest to those who work with geographic information but also to social scientists studying social, economic and political changes over time.

Biography:

Alex Chaucer is an INF PhD student with a primary specialization in Geographic Information Systems (GIS) and a secondary specialization in Information, Technology, and Learning (ITL). He has been in the program part time over a number of years. Alex currently works in the IT department at Skidmore College.

 

Andy Whitmore , Informatics

Full Information Product Pricing (FIPP): Information Policies to Expand the Share of Worker and Environment Friendly Trade in the NAFTA Region

Abstract:

This work focuses on the creation of a simulation model of one agricultural product - coffee grown in Mexico but roasted, brewed, and consumed in Canada and the United States -that realistically represents the embedded agricultural and economic systems and captures the dynamic interactions of this natural resource economy. The simulation model will be linked to a judgment task which will be performed to elicit user preference functions. The cues present in the judgment task scenarios will be elicited from a focus group session aimed at determining what pieces of product information are most useful to stakeholders in their market and policy choices. The objective functions emerging from the judgment task will be input into the simulation model where a set of optimal policies will be determined for each stakeholder through the maximization of the function. The purpose of this work is to create a means through which policy makers and consumers can identify an optimal set of policy options to guide their policy or purchasing decisions and to illustrate the importance of non-price product information in these choices. This research is part of a larger project undertaken by the North American Digital Government Working Group which is supported by the National Science Foundation, the National Council for Science and Technology (CONACYT, Mexico) as well as by institutions in Canada, Mexico, and the United States. The purpose of this broader research project is to explore a set of government-sponsored product labeling and information policies that may have the potential to significantly expand the share of fair wage and environmentally friendly products traded within the NAFTA region and thereby promote agricultural and economic sustainability.

Biography:

Andy Whitmore is an INF PhD student specializing in DAPS. Andy uses computer simulation to analyze policy problems in a range of fields including e-government, open source software and economics of information. Andy was previously an economist at the Consumer Price Index and an IT Auditor at Ernst & Young LLP. He has a BA in Economics from Cornell University and an MA in Information Systems from Johns Hopkins University.

 

Steve Lackey , Informatics

Alternative Transportation for Community College Students

Abstract:

Transportation has become an increasing burden on colleges, communities, and to the students themselves. Colleges often focus on Total Demand Management (TDM) strategies to discourage student parking via pricing disincentives. However, the lack of alternatives typically results in higher barriers to access for the most vulnerable members of the service population.

While Community Colleges enroll 40% of all college students annually, they typically serve a population spread over a 25-mile radius, forming a classic suburb-to-suburb commuting pattern that defies traditional mass transit. However, the greatest price sensitivity, greatest commuting burden, and lowest socioeconomic support fall on the community college student; financial difficulties are the second-greatest cause of attrition among students.

TDM often proposes ridesharing ("carpooling") as a means to reduce parking costs. However, traditional ridesharing has become limited to same-household use, and to a lesser extent where promoted by a common employer. Limitations on sharing personal schedules outside the household is seen as a major barrier to ridesharing outside the household, rather than the availability of potential ridesharers in the vicinity.

It is proposed in my research that an effective mechanism of selectively sharing Personal Information Management data will promote the formation and maintenance of rideshare groups. By selectively coordinating schedules and activities, rideshare opportunities can be planned beforehand via traditional "stationary" computing tools, while mobile computing tools (typically a "smart phone") may be used for contingencies such as schedule changes within the group. It is further anticipated that integrating an enhaced Personal Information Manager may help students manage their schedules and tasks to improve student outcomes.

Biography:

 

Tino DeMarco, Informatics

Complex Decision Making and Geographic Information Systems (GIS): The Residential Estate Market

Abstract:

It is often said that information may make the consumer purchase decision process easier. On the other hand, it is also stated that too much information may cause information overload and thus confusion. This dilemma may be most complicated in the world of complex purchase decision making where consumers may act irrationally while relying on multiple internal (experience etc.) and external (intermediaries, friends, experts, TV shows, etc.) information sources to help make a decision.

This paper seeks to develop a method by which geographic information systems (GIS) may replace the traditional “broker focused” or intermediated residential real estate purchase process and to determine the impact. We propose an experiment where residential home purchasers will forgo contacts with real estate brokers and instead concentrate on searching for a home relying on inputs from information systems.

An attempt will be made to measure the impact of these information systems primarily GIS mapping and overlay capabilities, on potential buyers while they move through the different stages of the purchase decision process. How will buyers search for and establish alternatives and make a final decision using these GIS without realtor inputs? Attention will focus on the post purchase stage of this decision process to gather input from buyers to learn how satisfied buyers are after the purchase using each process.

Biography:

Tino DeMarco is a 2nd year Ph.D. student in the University at Albany's College of Computing and Information Sciences. His primary interest is Information and Organization Environments (IOE) and his secondary interest is Geographic Information Systems (GIS). His research interests include Information and Communications Technologies (ICTs) in complex decision environments, GIS and behavioral economics. Tino currently works in the marketing department within the School of Business at the University at Albany, State University of New York. Prior to working at the university, Tino was a banker at Key Bank of New York and also served as an Adjutant General Officer working in Public Affairs at the New York Army National Guard. He is also an entrepreneur interested in property management, preservation and architecture. Tino holds a B.S. in Finance from Siena College and his M.B.A. in Marketing from the University at Albany, State University of New York.

 

Xiaowen Song, Informatics

Pilot Study on GIS Visualization

Abstract:

GIS visualization is a rising technology in GIS and transportation area, and it is still in the technology adoption phase. It improves communication between transportation agencies and public, and communication within the transportation project teams through advanced technologies such as computer graphic, animation and simulation. This research does literature review as a pilot study in the field of GIS visualization, summarizes the previous research work in this field, and tries to point out the gaps and problems in the framework as the direction for the future research.

Biography:

Xiaowen Song is the 2nd year student in Informatics at the University of Albany. Her primary specialization is Geographic Information Science (GIS), and her secondary specialization is Information Assurance (IA). Xiaowen has a Bachelor degree in Library Science of Anhui University (CHINA) and Master Degree in Management Information System of East China Normal University (CHINA). Currently, she is interested in the Geographic Information System Visualization technologies in transportation and urban planning area, and is doing her research in this area.

 




A Sample from Last Year's Conference!

image of a poster presentation
image of a poster presentation
image of a poster presentation